A progressive relaxation training method based on facial and body behavior perception

Through facial and body behavior perception technology, the Blaze Face and Blaze Pose algorithms are used to obtain feature values ​​and control the audio of progressive relaxation training, which solves the problem of lack of feedback and personalized guidance in traditional training, improves the efficiency and effect of relaxation training, and achieves the realization of A more accurate experience of full body relaxation.

CN118538367BActive Publication Date: 2024-11-19NANJING UNIV OF INFORMATION SCI & TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411003151.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-25
Publication Date
2024-11-19
Estimated Expiration
2044-07-25

AI Technical Summary

Technical Problem

Traditional progressive muscle relaxation training lacks effective feedback and personalized guidance, resulting in poor results in relaxation training.

Method used

Using computer vision technology based on facial and body behavior perception, facial and body key points information is obtained through Blaze Face and Blaze Pose algorithms, and feature values ​​are calculated to control the playback of progressive relaxation training audio, providing personalized feedback.

Benefits of technology

It improves the efficiency and effect of relaxation training, enhances the perception of tension and relaxation states in different parts of the body, provides a more precise and personalized relaxation experience, and promotes overall body relaxation and health management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118538367B_ABST
    Figure CN118538367B_ABST
Patent Text Reader

Abstract

The invention discloses a progressive relaxation training method based on facial and body behavior perception in the field of biological behavior recognition, comprising: playing pre-designed audio for progressive relaxation training; performing the following training operations during audio playing: acquiring image data of a person to be trained; wherein the image data comprises a facial image and a body image; acquiring facial key point information through a Blaze Face algorithm based on the facial image, wherein the facial key point information comprises eye key point information and mouth key point information; calculating eye opening and closing ratio according to the eye key point information, and calculating mouth opening and closing ratio according to the mouth key point information; acquiring body key point information through a Blaze Pose algorithm based on the body image, and calculating the head-shoulder angle; determining eye feature values, mouth feature values ​​and body feature values ​​according to the eye opening and closing ratio, the mouth opening and closing ratio and the head-shoulder angle; and controlling the playing of the training audio according to the eye feature values, the mouth feature values ​​and the body feature values.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of biological behavior recognition and emotional computing technology, and in particular to a progressive relaxation training method based on facial and body behavior perception. Background Art

[0002] Progressive muscle relaxation is a technique that helps reduce physical stress and tension. Its core concept is to achieve the effect of whole-body relaxation by consciously tensing and then relaxing muscle groups in various parts of the body. This technique has been widely used in areas such as relieving anxiety, improving sleep quality, treating headaches and chronic pain management. The practice of progressive muscle relaxation technique usually involves different parts of the body, generally starting from the feet and gradually moving up to the head, or vice versa. In this process, the subjects are first guided to tense each group of muscles for about 5 to 10 seconds, and then relax these muscles for 20 seconds. However, in the traditional progressive muscle relaxation process, practitioners are easily affected by human factors, lack feedback and personalized guidance, resulting in poor relaxation training results. Summary of the invention

[0003] The present invention provides a progressive relaxation training method based on facial and body behavior perception. The method utilizes computer vision technology to automatically identify facial expressions and body behaviors, and finally realizes progressive muscle relaxation of a trainee based on facial expression feature values ​​and body behavior feature values, thereby overcoming the problem of lack of effective feedback and personalized guidance in traditional progressive muscle relaxation training and improving the efficiency of relaxation training.

[0004] In a first aspect, the present invention provides a progressive relaxation training method based on facial and body behavior perception, comprising:

[0005] Play the audio pre-designed for progressive relaxation training; while the audio is playing, perform the following training operations:

[0006] Acquire image data of a person to be trained; wherein the image data includes a facial image and a body image;

[0007] Based on the facial image, the facial key point information of the trained person is obtained by using the Blaze Face algorithm, wherein the facial key point information includes eye key point information and mouth key point information; the eye opening and closing ratio is calculated according to the eye key point information, and the mouth opening and closing ratio is calculated according to the mouth key point information;

[0008] Based on the body image, obtaining body key point information of the trainee through the Blaze Pose algorithm, and calculating the head-shoulder angle according to the body key point information;

[0009] According to the ratio of eyes opening and closing, mouth opening and closing and head-shoulder angle, the eye feature value, mouth feature value and body feature value are determined;

[0010] The playing of the training audio is controlled according to the eye feature value, the mouth feature value and the body feature value.

[0011] In some embodiments, based on the facial image, obtaining facial key point information of the trained person through the Blaze Face algorithm includes:

[0012] The facial image of the person being trained is detected using the Blaze Face algorithm to obtain 12 key point information of the left and right eyes and 4 key point information of the mouth;

[0013] The 12 key points of the left eye and the right eye are evenly distributed around the eyes, and are recorded as:

[0014] ;

[0015] Among them, e1-e 12 Represents the 12 key points of both eyes, (x i ,y i ) represents the horizontal and vertical coordinates of the corresponding points in the image, and i ranges from 1 to 12.

[0016] The four key points of the mouth are located at the two corners of the mouth and the middle of the upper and lower lips, respectively, and are recorded as:

[0017] ;

[0018] Among them, m1-m4 represent the four key points of the mouth, (x i ,y i ) represents the horizontal and vertical coordinates of the corresponding points in the image, and i takes values ​​of 21-24.

[0019] In some embodiments, the binocular opening and closing ratio is calculated based on the binocular key point information. ,include:

[0020] ,

[0021] In the formula, the function Represents the calculation of the Euclidean distance between two coordinates.

[0022] In some embodiments, the mouth opening and closing ratio is calculated based on the mouth key point information. ,include:

[0023] ,

[0024] In the formula, the function Represents the calculation of the Euclidean distance between two coordinates.

[0025] In some embodiments, based on the body image, obtaining the body key point information of the trained person through the Blaze Pose algorithm includes:

[0026] The body image of the person being trained is detected by the Blaze Pose algorithm to obtain the four key points of the body; the four key points are located at the two corners of the mouth and two points symmetrically distributed on the shoulders, recorded as:

[0027] ,

[0028] Among them, b1-b4 represent the four key points of the body, (x i ,y i ) represents the horizontal and vertical coordinates of the corresponding points in the image, and i takes values ​​of 31-34.

[0029] In some embodiments, the head-shoulder angle is calculated based on the body key point information. ,include:

[0030] .

[0031] In some embodiments, the eye feature value, the mouth feature value, and the body feature value are determined according to the eye opening and closing ratio, the mouth opening and closing ratio, and the head-shoulder angle, including:

[0032] Eye feature values , is the opening and closing ratio of both eyes;

[0033] Mouth feature value , is the mouth opening and closing ratio;

[0034] Body shape characteristics , is the head-shoulder angle.

[0035] In some embodiments, the playing of audio pre-designed for progressive relaxation training includes:

[0036] Pre-designed in order Audio, represents the i-th audio, Indicates the number of times the i-th audio is played;

[0037] Play Audio , and make the current audio Number of plays The value of the current audio is increased by one. The number of times the audio is played meets the set frequency range requirement. If so, continue playing and execute the training operation. If not, stop playing the audio.

[0038] In some embodiments, controlling the playback of the training audio according to the eye feature value, the mouth feature value, and the body feature value includes:

[0039] Determine whether the eye feature value, mouth feature value and body feature value match the preset reference value. If they match and there is still audio that has not been played, play the next audio; if they do not match, replay the current audio , and make the current audio Number of plays The value of is increased by one.

[0040] In a second aspect, the present invention provides a computer-readable storage medium having a computer program / instruction stored thereon. When the computer program / instruction is executed by a processor, the steps of the progressive relaxation training method based on facial and body behavior perception described in the first aspect are implemented.

[0041] Beneficial effects: The present invention adopts a machine-to-human approach, uses computer vision technology to automatically identify the facial expression and body behavior of the organism, and combines progressive muscle relaxation training to enhance the subject's perception of the tension and relaxation of different parts of the body, thereby promoting full-body relaxation. The present invention further enhances the perception of tension and relaxation of different parts of the body through the progressive relaxation method based on facial and body behavior perception by combining computer vision technology, providing practitioners with a more accurate and personalized relaxation training experience, and effectively promoting full-body relaxation and health management. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 Schematic diagram of facial key points detected in an embodiment of the present invention;

[0043] Figure 2 Schematic diagram of key points of the body detected in an embodiment of the present invention;

[0044] Figure 3 Schematic diagram of the overall process of the progressive relaxation method based on facial and body behavior perception in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] Embodiment 1: This embodiment introduces a progressive relaxation training method based on facial and body behavior perception, comprising:

[0046] Play the audio pre-designed for progressive relaxation training; while the audio is playing, perform the following training operations:

[0047] Acquire image data of a person to be trained; wherein the image data includes a facial image and a body image;

[0048] Based on the facial image, the facial key point information of the trained person is obtained by using the Blaze Face algorithm, wherein the facial key point information includes eye key point information and mouth key point information; the eye opening and closing ratio is calculated according to the eye key point information, and the mouth opening and closing ratio is calculated according to the mouth key point information;

[0049] Based on the body image, obtaining body key point information of the trainee through the Blaze Pose algorithm, and calculating the head-shoulder angle according to the body key point information;

[0050] According to the ratio of eyes opening and closing, mouth opening and closing and head-shoulder angle, the eye feature value, mouth feature value and body feature value are determined;

[0051] The playing of the training audio is controlled according to the eye feature value, the mouth feature value and the body feature value.

[0052] The present invention can obtain the facial key point information and body key point information of a trainee through computer vision technology, and can obtain the eye opening and closing ratio, mouth opening and closing ratio and head-shoulder angle through these key point information; thereby, according to the eye opening and closing ratio, mouth opening and closing ratio and head-shoulder angle, eye feature values, mouth feature values ​​and body feature values ​​based on thresholds are respectively obtained, and then personalized progressive muscle relaxation training of the trainee is realized according to these feature values.

[0053] Embodiment 2: Based on Embodiment 1, this embodiment introduces a specific implementation process of a progressive relaxation training method based on facial and body behavior perception, such as Figure 3 As shown, specifically including the following:

[0054] 1. Acquisition of key point information

[0055] Based on the facial image, the facial key point information of the trained person is obtained by using the Blaze Face algorithm, including:

[0056] The facial image of the trained person is detected by the Blaze Face algorithm to obtain the 12 key points of the left and right eyes and the 4 key points of the mouth; Figure 1 As shown, the 12 key point information of the left eye and the right eye are evenly distributed around the eyes, recorded as:

[0057] ,

[0058] Among them, e1-e 12 Represents the 12 key points of both eyes, (x i ,yi ) represents the values ​​of the horizontal and vertical coordinates of the corresponding points in the image, and i ranges from 1 to 12;

[0059] The four key points of the mouth are located at the two corners of the mouth and the middle of the upper and lower lips, respectively, and are recorded as:

[0060] ;

[0061] Among them, m1-m4 represent the four key points of the mouth, (x i ,y i ) represents the values ​​of the horizontal and vertical coordinates of the corresponding points in the image, and i takes a value of 21-24. Thus, all the key point information for recognizing facial expressions is obtained.

[0062] Based on the body image, the key point information of the body of the trainee is obtained by using the Blaze Pose algorithm, including: detecting the body image of the trainee by using the Blaze Pose algorithm to obtain four key point information of the body; Figure 2 As shown, the four key points are located at the two corners of the mouth and two points symmetrically distributed on the shoulders, recorded as , where b1-b4 represent the four key points of the body, (x i ,y i ) represents the values ​​of the horizontal and vertical coordinates of the corresponding points in the image, and i takes a value of 31-34. Thus, all the key point information for body features is obtained.

[0063] Depend on Figure 1 and Figure 2 As shown in the figure, although the two key points b1 and b2 in the body are at the same position as the key points m1 and m2 in the mouth in the facial expression, the two pairs of coordinate points are detected using different models. The two models accept different images, and the latter can only return the key points of the face, while the former can return the key points of the whole body and both sides of the mouth.

[0064] 2. Calculate eigenvalues

[0065] Calculate the binocular opening and closing ratio: Calculate the binocular opening and closing ratio based on the binocular key point information , the specific calculation formula is as follows:

[0066]

[0067] In the formula, the function Represents the calculation of the Euclidean distance between two coordinates.

[0068] Threshold-based eye feature values:

[0069] ,

[0070] In the formula, is the eye feature value of the trained person, when =0, indicating that the trained person opens his eyes. =1, it means the person being trained has his eyes closed.

[0071] Calculate the mouth opening and closing ratio: Calculate the mouth opening and closing ratio based on the mouth key point information , the specific calculation formula is as follows:

[0072]

[0073] Threshold-based mouth feature values:

[0074] ,

[0075] In the formula, is the mouth feature value of the trained person, when =0, indicating that the person being trained opens his mouth. =1, it means the trainee keeps his mouth shut.

[0076] Calculate the head-shoulder angle: Calculate the head-shoulder angle based on the key point information of the body , the specific calculation formula is as follows:

[0077]

[0078] Threshold-based body feature values:

[0079] ,

[0080] In the formula, is the body feature value of the trained person, when =0, indicating that the trainee is in a relaxed posture. =1, it means the trainee is not relaxed.

[0081] 3. Control the playback of training audio according to the number of playbacks and each feature value

[0082] definition is the audio to be played, the number of audios is N, For instructions The number of times the audio is played, i is the order value of the currently played audio, that is, the i-th audio among N audios. , .

[0083] After determining the calculation and representative meaning of the eye feature value, mouth feature value and body feature value of the trainee, the execution status under each combination is given respectively. When the user opens his eyes or mouth during relaxation training or the head-shoulder angle is not within the range of 80 and the mouth, that is, in the five cases of Fe=0, Fm=0, Fb=0; or Fe=0, Fm=0, Fb=1; or Fe=0, Fm=1, Fb=1; or Fe=1, Fm=0, Fb=0; or Fe=1, Fm=1, Fb=0, it is considered that the user has not been fully relaxed. It will be played again. Only when Fe=Fm=Fb=1, it means that the user is fully relaxed, and then the next audio is played. The specific summary is as follows:

[0084] In some embodiments, the playing of audio pre-designed for progressive relaxation training includes:

[0085] Pre-designed in order Audio, represents the i-th audio, Indicates the number of times the i-th audio is played;

[0086] Play Audio , and make the current audio Number of plays The value of the current audio is increased by one. The number of times the audio is played meets the set frequency range requirement. If so, continue playing and execute the training operation. If not, stop playing the audio.

[0087] In some embodiments, controlling the playback of the training audio according to the eye feature value, the mouth feature value, and the body feature value includes:

[0088] Determine whether the eye feature value, mouth feature value and body feature value match the preset reference value. If they match and there is still audio that has not been played, play the next audio; if they do not match, replay the current audio , and make the current audio Number of plays The value of is increased by one.

[0089] In this example, preferably, Figure 3 As shown, the method for enabling the trainee to achieve personalized progressive muscle relaxation training comprises:

[0090] S1, Initialization , ,in , , , , ;

[0091] S2, play audio , ;

[0092] S3, Get The value of

[0093] S4, if , go to step S5, otherwise go to step S9;

[0094] S5, if The value of is 1, then , go to step S6; otherwise, return to step S2;

[0095] S6, if , play the next audio Pi and return to step S5; otherwise, go to step S8;

[0096] S7, play prompt: relaxation training is completed, and go to step S9;

[0097] S8, play a prompt: it is not suitable to do relaxation training at this time, and go to step S9;

[0098] S9, training ends.

[0099] The above method steps determine whether to play audio and whether to play the next audio by judging the status of the eye feature value, mouth feature value and body feature value; and judge whether the user is suitable for relaxation training by judging the number of times the current audio is played. When the number of times the training is greater than 4, the training is judged to be finished because when the path reaches this point, it indicates that the same audio has been played 4 times, and the user has not yet entered a state of relaxation, so a prompt is given. The invention can adjust the training program in real time according to the training state of the trainee to adapt to the actual situation and progress of the trainee, thereby realizing personalized relaxation training and improving training efficiency.

[0100] Embodiment 3: This embodiment introduces a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the progressive relaxation training method based on facial and body behavior perception as introduced in Embodiment 1 or 2 are implemented.

[0101] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0102] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0103] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0104] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0105] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.

Claims

1. A progressive relaxation training method based on facial and body behavior perception, characterized in that: include: Play pre-designed audio for progressive relaxation training; During audio playback, perform the following training operations: Acquire image data of a person to be trained; wherein the image data includes a facial image and a body image; Based on the facial image, the facial key point information of the trained person is obtained by using the Blaze Face algorithm, wherein the facial key point information includes eye key point information and mouth key point information; the eye opening and closing ratio is calculated according to the eye key point information, and the mouth opening and closing ratio is calculated according to the mouth key point information; Based on the body image, obtaining body key point information of the trainee through the Blaze Pose algorithm, and calculating the head-shoulder angle according to the body key point information; According to the ratio of eyes opening and closing, mouth opening and closing and head-shoulder angle, the eye feature value, mouth feature value and body feature value are determined; Controlling the playback of the training audio according to the eye feature value, the mouth feature value, and the body feature value; Wherein, based on the facial image, obtaining facial key point information of the trained person through the Blaze Face algorithm includes: The facial image of the trained person is detected by the Blaze Face algorithm to obtain the 12 key points of the left and right eyes and the 4 key points of the mouth; The 12 key points of the left eye and the right eye include key points e1-e6 that are evenly distributed along the periphery of the first eye, and key points e7-e8 that are evenly distributed along the periphery of the second eye. 12 , where key points e1 and e4 are located at the two corners of the first eye, and key points e7 and e 10 Located at the two corners of the second eye, respectively, recorded as: ; Among them, e1-e 12 Represents the 12 key points of both eyes, (x i ,y i ) represents the horizontal and vertical coordinates of the corresponding points in the image, and i ranges from 1 to 12. The four key points of the mouth are located at the two corners of the mouth and the middle of the upper and lower lips, respectively, and are recorded as: ; Among them, m1-m4 represent the four key points of the mouth, (x i ,y i ) represents the values ​​of the horizontal and vertical coordinates of the corresponding points in the image, and the value of i is 21-24; Calculate the binocular opening and closing ratio according to the binocular key point information ,include: , In the formula, the function Indicates calculating the Euclidean distance between two coordinates; Calculate the mouth opening and closing ratio according to the mouth key point information ,include: , Based on the body image, the key point information of the body of the person being trained is obtained through the Blaze Pose algorithm, including: The body image of the person being trained is detected by the Blaze Pose algorithm to obtain the four key points of the body; the four key points are located at the two corners of the mouth and two points symmetrically distributed on the shoulders, recorded as: , Among them, b1-b4 represent the four key points of the body, (x i ,y i ) represents the values ​​of the horizontal and vertical coordinates of the corresponding points in the image, and i takes values ​​of 31-34; Calculate the head-shoulder angle based on the body key point information ,include: ; According to the ratio of eyes opening and closing, mouth opening and closing and head-shoulder angle, the eye feature value, mouth feature value and body feature value are determined, including: Eye feature values , is the opening and closing ratio of both eyes; Mouth feature value , is the mouth opening and closing ratio; Body shape characteristics , is the head-shoulder angle; The playing of audio pre-designed for progressive relaxation training includes: Pre-designed in order Audio, represents the i-th audio, Indicates the number of times the i-th audio is played; Play Audio , and make the current audio Number of plays The value of the current audio is increased by one. whether the number of times of playing the audio meets the set frequency range requirement, if so, continue playing and perform the training operation, if not, stop playing the audio; Controlling the playback of the training audio according to the eye feature value, the mouth feature value, and the body feature value, including: Determine whether the eye feature value, mouth feature value and body feature value meet F e =F m =F b =1, if satisfied and there is still audio to be played, play the next audio; if not satisfied, replay the current audio , and make the current audio Number of plays The value of is increased by one.

2. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the steps of the progressive relaxation training method based on facial and body behavior perception described in claim 1 are implemented.